Stable diffusion model (Keras-CV)
Budget: $50 – $100 USD
in this project, you need to use fine-tuning stable diffusion from keras-cv library. you can find it here: https://keras.io/examples/generative/finetune_stable_diffusion/.
1) choose any dataset sample you prefer but it should be image-caption format and it follows the above link requirements.
2) fine-tune epoch, guidance score, and hyperparameters to find the best result (show report of experimentation results)
3) use and show Kernel Inception Distance (KID), and Fréchet Inception Distance (FID) metrics for training and testing. (show report of the results)
5) show use of different tokenization techniques/data augmentation is a plus
**Important**
*Code should be clean, clear, and commented in good comprehensive way.
*Code should work with Colab Pro+
1) choose any dataset sample you prefer but it should be image-caption format and it follows the above link requirements.
2) fine-tune epoch, guidance score, and hyperparameters to find the best result (show report of experimentation results)
3) use and show Kernel Inception Distance (KID), and Fréchet Inception Distance (FID) metrics for training and testing. (show report of the results)
5) show use of different tokenization techniques/data augmentation is a plus
**Important**
*Code should be clean, clear, and commented in good comprehensive way.
*Code should work with Colab Pro+